Compare AI models

Specifications
Parameters
—2.4T
Context window
128k1M
Benchmarks
MMMU-Pro
43.8%82.3%
Overview
CompanyMistralQwen
Release dateAug 12 2025Aug 3 2026
AccessClosedClosed
Model detailsView modelView model

Frequently asked questions

Qwen3.8-Max leads Mistral Medium 3.1 on 1 of the 1 benchmark they both report (MMMU-Pro). Mistral Medium 3.1 shipped 356 days before Qwen3.8-Max, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (Mistral Medium 3.1) vs 1M (Qwen3.8-Max).

On MMMU-Pro, Qwen3.8-Max leads at 82.3% vs Mistral Medium 3.1 at 43.8%.